July 19, 2009
Natural gas and oil prices
Since the start of the year, the price of crude oil has risen about 40% while the price of natural gas has fallen by about 40%. Can that divergence be maintained?
A barrel of oil has about six times the energy content of a thousand cubic feet of natural gas. The graph below compares the dollar price of a barrel of oil with the oil-equivalent cost of natural gas, calculated by multiplying the price (in $/1000 cu ft) by six. The two prices tended to move together in the early part of the decade, but have diverged significantly over the last few years, with natural gas today selling for 1/3 the price of oil in terms of BTU content.
Does this mean that oil prices will go down, natural gas prices will go up, or both? Let Δot denote the monthly percent change in in oil prices (technically, the change in the natural logarithm) and zt the percentage gap in cost (technically, zt = ln(ot/6gt)). If you use a regression to try to predict oil prices from their own lagged values and the lagged oil-gas cost gap, a positive gap such as we have at the moment does tend to tug down future oil prices slightly, though the coefficient is not statistically significant. Here are the regression coefficients, with standard errors in parentheses:
On the other hand, the cost gap does seem to help significantly to predict where natural gas prices might go. With the gap currently at zt = 1.13, the historical regression below might lead you to expect natural gas prices to climb by 10% a month (0.103 x 1.13 = 0.116) until the gap is closed.
But it's hard to see that happening in the face of news like this:
Natural gas prices have been pummeled by over-supply and weak demand, cutting the national drill rig count in half in the past year. Meanwhile, monstrous initial production rates in non-traditional shale plays have added to price pressure, as producers fight to hold the terms of their expensive leases in boomtowns....
Now, British Columbia's Horn River Basin might be added to the list, with the help of behemoth ExxonMobil (XOM) reportedly coming up with initial rates on early test wells to the tune of 16- to 18-million cubic feet a day.... That is double the rate of a really good well and in line with the Haynesville, in Louisiana, which may hold some 250 trillion cubic feet of recoverable national gas, enough to satisfy domestic demand for a decade.
So maybe that negative coefficient in the first regression will do some work after all, tugging oil prices down. In the mean time, trends like this are worth encouraging:
Mass transit buses have operated on natural gas for years and in other countries for decades. According to NGVAmerica, some 125 transit agencies are now operating over 10,000 natural gas transit buses on American highways. This trend is expanding, with over 20 percent of all new bus orders offering natural gas propulsion.
Posted by James Hamilton at July 19, 2009 06:35 AMdigg this | reddit
So, what is the proper role for the Federal government in this?
What can we learn about a governmental role by reviewing the history of Canada's control and ownership of natural resources?
What is a good middle ground between the U.S. and Canada practice? Is any middle ground possible or available?
How do we keep development of new sources going during period of low prices and restrain the excesses during periods of high prices?
Posted by: RofermerRay at July 19, 2009 08:50 AM
How about a comparison of nat gas and coal? Coal is of course hard to compare since there are greater variations in the energy content among the types, but you could try it with a specific coal over the years, i.e. Powder River basin coal. See how those compare.
Another interesting chart might relate nat gas prices to nitrogen fertilizer prices and/or demand, or some other heavy user of nat gas.
It's kind of hard to compare nat gas, oil, and coal to each other apples to apples, since each has different uses and extra demand (or a smaller relative drop in demand) in any of the specialties (electricity, chemicals, transportation, etc), or innovative extraction methods (shale gas, secondary and tertiary recovery) can create large divergences in price, at least in the short run.
Posted by: Eric at July 19, 2009 10:21 AM
My understanding is the biggest factor keeping the markets distinct is that natural gas is harder to store and transport. So there are often significant price differentials for gas. In the Middle East, it's very cheap, in Western Europe, somewhat dear.
Oil is therefore also easier to speculate in, which I think is occurring now and accounts for some of the differential.
I think the recent rise in commodities is mostly based on a presumption that demand will pick up later, not that there is any real evidence it has already.
Here's a great illustration via metals. These are the charts for copper and aluminum, both used in all sorts of manufacturing and construction.
Take a look at the 6 month graph of LME stocks at the bottom of each page:
Copper's inventory has been falling for since March, and aluminum has been rising just as fast.
Commodity speculation through storage is occurring in oil, cooper and some others, but not aluminum and gas.
I believe we are in a mini-commodity bubble.
Posted by: Bob_in_MA at July 19, 2009 04:25 PM
Why is z(t) = ln(o(t)/6*g(t))? Specifically, what is the "six" for?
Posted by: ebh at July 20, 2009 05:30 AM
We are long nat gas due to success with horizontal drilling and hydrofracking.
We are short on oil due to...well, peak oil, plus or minus. Non-OPEC production peaked in May 2005. OPEC has added some capacity, but not too much. So, overall, we're near net flat in production capacity, with very significant pricing power returning to OPEC, probably this fall.
Nat gas futures are running at about $6, oil at around $75, which would give a 12:1 ratio, about 2x btu parity. I think that's about right, suggesting we should be making a whole-hearted push to nat gas for transportation fuel.
Government policy would best be directed to network effects, ie, subsidizing the standardization and distribution of natural gas in filling stations. This would enable the average consumer to have a real choice between gasoline and nat gas powered vehicles.
Posted by: Steve Kopits at July 20, 2009 05:51 AM
With very little capability for inter-fuel substitution, we're unlikely to return to oil-gas price linkage anytime soon. For gas, the supply curve is moving out faster than the demand, for oil, less so. As was mentioned above, the important substitute for gas is coal in power generation, and because of reduuced power demand, the spot coal price is getting crushed. So, with cheap coal and abundant, there's nothing to pull up the price of gas in the near term. A reduction in the amount of money spent on reservior stimulation in shale gas might lead to a tailing-off in production and a price move to $6 or $7 for gas, but that likely won't kick in for half a year or so.
Posted by: bartman at July 20, 2009 06:21 AM
Dr. H - It seems logical to me that natural gas should sell at a discount to oil relative to $/BTU, since it more difficult to transport and store, and has a lower energy density. In most cases, gas is not easily substituted for gasoline or diesel as a transport fuel, due to limitations in distribution and high conversion costs on engines and storage tanks.
The current wide divergence between oil and gas prices owe more to rampent speculation in oil markets than to common sense. My gut reaction is that investors have been looking towards tangible liquid assets rather than dicey Sovereign bonds whose value can quickly erode via inflation, foreign exchange valuation, or rising bond interest rates. The recent doubling in oil prices since January 2009 are more due to
the herd stampeding, and contango, than to common sense. There are still a lot of suckers that believe there will be a robust recovery of western economies within the next year.
Posted by: MarkS at July 20, 2009 07:22 AM
ebh: Without the 6, all the coefficients in the regression would be identical except for the constant term. I chose to include the 6 because I thought if I did so, people could follow the argument more quickly without having to wade through the above point.
Posted by: JDH at July 20, 2009 07:24 AM
The current wide divergence between oil and gas prices owe more to rampent speculation in oil markets than to common sense.
I know that the above notion is an article of faith amongst the tinfoil-hat brigade, but most people that actually work in the energy business understand that the fundamentals make perfect sense, in as much as human expectations can be said to "make sense". They made sense last July, they made sense last December, and they do now. Steep demand curves coupled with steep supply curves make for large price moves in small time intervals. I understand how this can seem mysterious to a layman, but if you understand the market structure, it isn't all that strange.
Posted by: bartman at July 20, 2009 07:58 AM
An anecdote -- I just moved to Northern Colorado and spoke with someone who is part-owner of natural gas drilling rigs in Wyoming. He said natural gas prices are so low that their drilling operations are currently idle, although they will likely start up again before winter to help keep the equipment in good working order. He was very aware of the price gap and even considered "playing the spread" in futures.
Posted by: JE at July 20, 2009 09:36 AM
One word: Gas to Liquids (GTL).
GTL diesel is... beautiful. Literally (it's as clear as vodka, unlike regular diesel).
GTL plants in the middle east are the best way to get that "stranded gas" to where it needs to go. And no need to convert busses and trucks to run on nat gas. GTL diesel runs in any diesel as-is.
Posted by: Buzzcut at July 20, 2009 11:53 AM
“This is the third straight month of a gain in leaders and suggests that, along with other economic evidence, the U.S. recession might have ended” in the second quarter, Kenneth Kim, an economist at Stone & McCarthy Research Associates in Princeton, New Jersey, wrote in a note to clients.
Posted by: Steve Kopits at July 20, 2009 12:01 PM
Buzzcut: One word: Gas to Liquids (GTL).
This report from the Dept. of Energy indicates that GTL is extremely capital intensive, costing $25,000 to $45,000 per daily barrel. We currently use about 4.3 million barrels of diesel per day. We use about twice that in gasoline. You can do the math on capital costs.
They also state that GTL is extremely sensitive to gas prices. They say GTL is viable when oil is $25 and gas is fifty cents to a dollar per million BTU. Gas is currently about $4 per BTU, so GTL might barely work if oil is $100 to $200 per barrel (assuming the price of gas doesn't also go up). Anyone want to make a trillion dollar bet?
In other words, GTL is only economically viable for stranded gas resources, that is, gas that is extremely cheap because it isn't near a pipeline.
Posted by: Anonymous at July 20, 2009 12:52 PM
My understanding is the recent (relative) collapse in ng prices is due to both a domestic and international excess supply of ng. This is likely to continue through the end of the year at least. A spike in US ng prices could come about from disruptions during the hurricane season.
Check EIA's latest NG Monthly report. You can see that inventories are well above norms. In previous work I have found periods of slackness or tightness in inventories provide substantial explanatory power in the monthly change in HH prices.
Posted by: fred at July 20, 2009 01:36 PM
This price divergence mainly demonstrates the power of China on the global oil market. US natural gas has become relatively less valuable because China can't use it.
Posted by: Tom at July 20, 2009 03:17 PM
1) Seasonal effects make the graph a bit overly dramatic. Looking at summertime prices compares oil at the top of its seasonal cycle to natgas at the bottom of its cycle. Don't get me wrong though, I am not claiming that seasonal effects are anything like the whole story, and I see that the regression takes account of seasonal effects. I am just saying that the graph is slightly misleading.
2) Residual fuel oil prices make a useful comparison, since some power plants can fuel-switch between natgas and resid. Looking at EIA data I am getting a resid price of $34 - $37 per crude bbl heat equivalent - or roughly double the natgas price - from Nov 2008 through Apr 2009. That makes me think this gap is going to be transitory.
Posted by: Strat at July 20, 2009 05:56 PM
GTL is dumb if you have a local customer of that diesel. It is just smarter to just switch part of the transportation diesel users over to gas. Diesel trains and urban mass transport are the first candidates to do that.
Posted by: Charly at July 20, 2009 06:24 PM
Professor Hamilton...to advance my skills, I would like to reproduce your regression. Looks like I can get to the data...did you use an AR model? What stat package did you use? Many thanks for everything.
Posted by: Steve Bannister at July 21, 2009 09:55 AM
Steve Bannister: These are simple OLS regressions, and you should get the identical answer regardless of the package you use. I did these on RATS.
Posted by: JDH at July 21, 2009 10:12 AM
In other words, GTL is only economically viable for stranded gas resources, that is, gas that is extremely cheap because it isn't near a pipeline.
Right. That's why all the GTL plants are being built in the Emirates. There's a ton of stranded gas there, and GTL is the best way to deal with it.
And with global demand for diesel being so high, GTL diesel is the perfect product for stranded NG. If NG were used to make, say GTL gasoline, it would be a much less desireable product.
Charly, you can't "convert" a diesel engine to run on nat gas. You convert a diesel vehicle to NG by ripping out the diesel engine and replacing it with a spark ignition one. While not impossible, this is not an easy conversion. Better to just buy a NG powered vehicle in the first place.
The thrifty among you might look into what it takes to convert your personal vehicle to NG. Or look for an ex-municipal vehicle that is already NG powered.
Posted by: Buzzcut at July 22, 2009 08:14 AM
"The current wide divergence between oil and gas prices owe more to rampant speculation in oil markets than to common sense."
You immediately got my attention, when you characterized this sentence as "an article of faith amongst the tinfoil-hat brigade". YIKES... I thought I was too young to be a John Bircher.
I'm well aware of the historical rising oil demand curve, and flattened oil supply curve. However, I am also aware that the worldwide recession and financial crisis is putting a huge dent into demand, and current production is out-pacing consumption. I believe that it makes perfectly good sense to buy oil wells as a longer term investment when the world economy recovers in somewhere between 2013 and 2015 (when most of the bad RE debt will have been realized)... To me, it doesn't make sense to buy spot oil and store it in the hope that prices will rise fast enough in the short-term to pay for the storage and the capital allocation.
Perhaps our argument is concerned more about how the general world economy will be doing in the next three or four years than about oil process per-se. I expect that no bubble will emerge, and greater efforts will be made in Europe and the US to pay down debt and realize losses than to pumping up the consumer economy. I would speculate that you have a far more optimistic prognostication of the near-future than I do.
To me, the banking and finance industries remain fragile. Another black-swan can crush any gains the markets have made in the last four months almost over-night.
Posted by: MarkS at July 22, 2009 02:09 PM
But a GTL plant isn´t build instantly either so the change can be done with the natural rate of replacement.
ps. I assumed that diesel trains were turbines but the seem to be all diesel engined.
Posted by: Charly at July 22, 2009 05:56 PM
I respect and occasionally check in on JDH's blog posts here at Econbrowser, but I have to question the use of regressions to draw any conclusions about such complex market phenomena, such as "maybe that negative coefficient in the first regression will do some work after all." I do get the impression that that is how you made your bones, by doing such math better than the other mathematically illiterate economists, so hopefully you also have the insight to realize that it's basically astrology with a patina of mathematics. With your reputation, maybe you could lead the economics profession away from this disturbing trend, which thankfully other social scientists are too mathematically illiterate to ape. ;)
Posted by: Ajay at August 1, 2009 10:14 PM
The industry has been trying to push the price of peto to be equivalent in BTU properties.
The price of electricity for instance, at least in Alberta is pinned to the BTU capability.
A broader study would be appreciated
Posted by: cyberclark at August 24, 2009 09:47 AM
Interesting question about price divergence, but I have a bigger question. Last year, I saw your terrific presentation at AAEA and enjoyed your blog coverage of the question: what fossil fuel price level is required to promote substantial conservation (more public transport and fewer automobile miles) and alternative energy (wind and solar)? I've been confused by price fluctuations in the past year. Are prices currently high enough for good environmental progress?
Posted by: Parke at August 28, 2009 06:05 AM